The Changing Face of Quality Management

The panel first looked at the future role of quality management in life science. James Man of Syneos Health suggested that the activity could now be more embedded within everyday operations and improved incrementally, a vision that Peter Brandstetter of Accenture agreed with.

“I think companies will reach a point where they don't need quality management or people working in quality,” he said. “When everything is digital and automated, quality management loses its relevance.”

On the ground, though, the situation looks rather more complex. Merck’s Heiner Niessen said he believed quality demands are actually increasing, as more parameters are measured and the responsibility for quality extends along the supply chain.

Yet this, in turn, increases the urgency to address the manual burden of quality monitoring and management. “There’s an opportunity here,” Brandstetter said. “The more we learn about the manufacturing process from all of the data being collected – and about the context – the better we can predict it, rather than checking retrospectively.”

It’s here that smart use of technology offers considerable potential – for example in managing information from different sources. As Merck’s Niessen put it: “With improving digitalisation, it’s much easier to capture and track this information, right out into the real world.”

The idea of smarter technology and automation can cause consternation among teams – who may worry about being sidelined, or about whether they can trust the technology sufficiently to let it take over from human oversight.

But Syneos Health’s James Man said that it is still people who make the decisions and promote change – however, if those decisions can be informed by rich data, the outcomes are likely to be better.

It might be better to involve quality experts more, he suggested. “Right now, there isn't typically a role of Chief Quality Officer – but perhaps there should be,” he said. “If companies want to be more pre-emptive, and for compliance to add value to the business, there will need to be structural changes – beginning with representation.”

At Merck, machine intelligence is already starting to play a role in quality management, although this is currently only in limited areas such as image analysis. Although AI has yet to be applied to ‘real decision-making’, Niessen suggested that the quality considerations would be similar to those already in place.

“Either you train the human being, or you train the algorithm, so if it doesn't work, the measures to take will be the same,” he noted. The critical requirement, though, would be that the algorithm’s own decision-making process would be transparent.

Syneos’s Man suggested that AI/ML might offer potential in the context of individualised/personalised medicine and quality management – to keep track of oversight and do this more cheaply, and less manually: along the lines of a digital twin perhaps.

“If AI can be applied as a learning tool, I do see potential for collectively improving capability by identifying near misses and so on,” he said.

Niessen noted that when very low quantities of a product are involved, the need for greater efficiency with quality management grows, given that there will be as many quality control measures as there are personalised products.

“Do companies just have to accept that quality costs will rise as data and parameters increase, or should they become better at reducing effort and containing costs?” asked James Kelleher at Generis.

There is enormous pressure on the industry to reduce the cost of medical drugs and companies do need to contain costs wherever possible, the panel agreed – especially with the trend towards personalised medicine. Also, from a health insurance perspective, as outcome-based reimbursement becomes more established, the cost of quality does become a factor, Accenture’s Brandstetter added.

So, what of the scope for smarter quality in delivering tangible business benefits – for example, driving insights for future products, preventing recalls, ensuring the supply chain delivers as expected, and so on?

“Lots of companies recognise that quality is an underutilised competitive lever,” James Man acknowledged. “People are cautious about investing currently,” he said. “There’s certainly more that can be done.”

To finish up the debate, Generis’s Kelleher asked the panel what definitive steps companies might take between now and 2025 to move closer to their goals.

Merck’s Niessen pointed to the need to connect individual quality systems across the value chain to enable seamless data transfer. “So, you would have your CAPA system, your RIM system, your supplier RIM system all acting more or less as one system,” he suggested. Merck sees a big advantage here, he said, and has initiatives in place to drive standard data exchange formats for exchanging quality information at a system level.

Accenture’s Brandstetter concurred that there needs to be more structured data that are exchangeable across company borders. Blockchain could help here, he suggested – enabling a trusted chain of data.

Regarding the role of static ‘documents’, the panel were undecided. James Man thought that their role would not diminish. “The way we access and interact with documents is here to stay,” he said. “But we should connect quality systems, and a Chief Quality Officer function will be important. Embedding quality people in the key R&D teams will happen. Achieve that, and you might be piloting more real-time data exchange with the regulators by 2025.”